Machine Learning Hot Seat Campaign (Machine Learning for RTW and M2TW) WIP
Hey guys at TWCenter.Net,
I am an amateur programmer and a current student at a data science bootcamp. I’ve derived a way to play against machine learning AI in both campaign and in battle, but it will be tedious and slow, unless I find a better solution via quicksaving, loading, quicksaving, so on and so forth.
By using the hotseat campaign mechanic, users can play against machine learning AI easily and quickly. The save file for the hotseat campaign called Quicksave.sav, can be used to readily have a trained ML agent play Rome Total War, save the Quicksave.sav file and the user can load the Quicksave.sav file. Note, the player cannot play in battles against the Machine Learning AI, so be careful how you play. There could theoretically be a function via a script for the game to by default, as per every attack by the AI, the game could stop, quick save, and the player could fight the battle.
That is a little more tricky, but I think doable with some neat scripting tricks. The Machine Learning AI will be fed videos of my gameplay footage per respective total war game and even better high quality footage later on to train on. This will be one of my projects I will undertake later on down the road. It is an idea that is a good loophole workaround.
Ask me any questions and if anyone wants to work on this project for the next two months, please email me at

john.lasheras@gmail.com or call me at 551-482-6232.

Maybe we can get good jobs out of this.

Sincerely,
John Lasheras